Pancreatic cancer places a significant burden on society and has no effective early detection method. A glyco-microarray approach will be used to search for early detection biomarkers of pancreatic cancers in human plasma. We will use multi-dimensional liquid phase fractionation of intact N-linked plasma glycoproteins previously isolated by lectin affinity columns. The multi-dimensional fractionation will involve nonporous chromatography to separate the glycoproteins and liquid capillary isoelectric focusing to separate protein isoforms, thus providing a means to collect isolated glycoforms in liquid phase for further analysis. UV absorption detection will allow profiling of changes between cancer versus control. Proteins of interest will be identified by mass spectrometry. These fractions will be spotted on nitrocellulose-coated microscope slides to produce a natural glycoprotein microarray, and will be interrogated by various fluorescently-labeled lectins to probe each microarray spot for the presence of different glycan moieties. Patients with pancreatic adenocarcinoma, pancreatic mucinous cystic neoplasms (MCNs) and intraductal papillary mucinous neoplasms (IPMNs), chronic pancreatitis. Type II diabetes for 10 or more year, and normals patients will serve as the disease categories of interest. Plasma from 30 participants per category will be analyzed to search for patterns that can discriminate patients with MCNs/IPMNS from the other disease categories. Glycoproteins that reveal such changes will be analyzed by QIT-TOF (MALDI-MSn) mass spectrometry to examine the detailed changes in glycan structure that may serve as biomarkers. Once the potential bio-markers are identified, high throughput antibody arrays will be used to establish information necessary to plan a validation study. This will include initial analytical validation to define the within and between indi-vidual varaibility using 30 participants per disease cateogry. Next, preliminary decision analysis will be per-fonned on an open label set of assays from 50 participants per disease category. Finally, a blinded set of assays will be done on 95 particiants per disease category. This systematic approach to examing the analytic characteristics of the assay will provide information requhed to plan a valdiation of these early detection biomarkers for pancreatic cancer. It is envisioned that these biomarkers could be used for early detection among high risk groups such a smoker, patients with long term Type II diabetes or chronic pancreatitis.